Genome-wide transcriptomic analysis of a desert willow, Salix psammophila, reveals the function of hub genes SpMDP1 and SpWRKY33 in drought tolerance
Drought is a major environmental constraint to plant growth, development and productivity. Compared with most willows that are generally susceptible to drought, the desert willow Salix psammophila has extraordinary adaptation to drought stress. However, its molecular basis of drought tolerance is still largely unknown.
During polyethylene glycol 6000 (PEG 6000)-simulated drought stress, we found that the osmotic adjustment substances were accumulated and the antioxidant enzyme activities were enhanced in S. psammophila roots. A total of 8172 differentially expressed genes were identified in roots of S. psammophila through RNA-Sequencing. Based on K-means clustering, their expression patterns were classified into nine clusters, which were enriched in several stress-related processes including transcriptional regulation, response to various stresses, cell death, etc. Moreover, 672 transcription factors from 45 gene families were differentially expressed under drought stress. Furthermore, a weighted gene co-expression network was constructed, and eight genes were identified as hub genes. We demonstrated the function of two hub genes, magnesium-dependent phosphatase 1 (SpMDP1) and SpWRKY33, through overexpression in Arabidopsis thaliana. Overexpression of the two hub genes enhanced the drought tolerance in transgenic plants, suggesting that the identification of candidate drought tolerance genes in this study was highly efficient and credible.
Our study analyzed the physiological and molecular responses to drought stress in S. psammophila, and these results contribute to dissect the mechanism of drought tolerance of S. psammophila and facilitate identification of critical genes involved in drought tolerance for willow breeding.
KeywordsSalix psammophila Drought Co-expression network Hub gene Transgene SpMDP1 SpWRKY33
Cauliflower mosaic virus
Differentially expressed genes
False discovery rate
Kyoto encyclopedia of genes and genomes
Late embryogenesis abundant hydroxyproline-rich glycoprotein
Magnesium-dependent phosphatase 1
Modifier of rudimentary
Quantitative real-time PCR
Quantitative trait loci
RING domain ligase2
Receptor-like cytoplasmic kinase
Relative water content
SIGMA FACTOR-INTERACTING PROTEIN1
Ubiquitin-conjugating enzyme E2
Weighted gene co-expression network analysis
Forest ecosystems account for ~ 30% of the land surface (covering ~ 42 million km2), providing many benefits to natural systems and humankind from ecological, economic, social, and aesthetic services [1, 2]. Drought is a major environmental constraint to plant growth, development and productivity. Notably, owing to the long-life spans of forest tree and difficulty for implementing irrigation on a large scale, drought greatly reduces tree productivity and survival across many forest ecosystems . The research of current climate models has predicted increased drought episodes due to the long-term effects of global warming . Therefore, expediting dissection of drought tolerance mechanism and improving drought adaptation of forest trees are urgent needed for arid land utilization, environmental sustainability and improvement of economic benefits.
To accommodate drought stress, plants have evolved sophisticated mechanisms to perceive external signal and coordinate metabolic pathways and morphological traits by modulating the expression of genes . Currently, many drought-inducible genes have been identified through transcriptome analysis in Arabidopsis [6, 7, 8], rice [9, 10, 11], maize , and other species. These genes can be classified into two classes: functional genes and regulatory genes . Overexpression of some drought-responsive genes (e.g., ethylene response factor 48, calcium-dependent protein kinase 4, NAC5, MYB5, etc.) has demonstrably enhanced drought tolerance in transgenic plants [13, 14, 15, 16]. Nevertheless, large variations occur in transcriptional responses to drought among different plant species [17, 18]. Most researches of drought responses mechanisms have concentrated in crops or model species. Research into forest trees is still at an initial stage except several species, such as Populus, Pinus, Eucalyptus and Quercus [19, 20]. Thus, studies on the species-specific molecular mechanisms of drought response are essential for molecular breeding for drought tolerance in forest trees.
Willows (Salix; Salicaceae) are a very diverse group of dioecious catkin-bearing trees and shrubs, which were the research focus as a potential source of sustainable and renewable biomass for bioenergy, biofuels and bioproducts [21, 22]. However, willows are generally susceptible to drought stress, so they commonly grow in alluvial or riparian habitats . In recent years, willow regulatory responses to drought at morphological, physiological, genomic and transcriptional levels have been examined in some studies, and some loci or candidate genes associated with drought responses were identified. For example, quantitative trait loci (QTLs) associated with drought tolerance for Salix dasyclados × S. viminalis hybrid in two water regimes (normal and drought-treated) have been identified, and each QTL explained 8 to 29% of the phenotypic variation . Fifteen willow genotypes are assessed the impact on the growth and leaf traits under permanent and temporary water stresses, and great variation in drought response among genotypes has been found . Two willow genotypes that originated from a cross between S. viminalis and S. viminalis × S. schwerinii and displayed different responses to drought have been performed phenotypic and transcriptional investigation under drought stress, and a set of candidate genes associated with drought response have been identified . Despite these advances, the comprehensive molecular mechanisms of drought resistance in willows remain poorly elucidated, and functional validation for detected candidate genes has not been carried out.
In willows, S. psammophila is an important shrub biomass willow that distributed in arid and semi-arid desert regions. It can well adapt to multiple abiotic stresses including drought, extreme temperature, sandstorm, and so on . So, it is used to prevent wind erosion and control desertification, playing vital roles in local vegetation rehabilitation . Furthermore, it is identified as a promising biomass feedstock [28, 29]. Drought tolerance plants that can maintain turgor and continue metabolism in cells even at low water potential are identified as potential sources for studying the drought tolerance mechanisms and exploring key regulatory genes for breeding . However, the drought response mechanism of S. psammophila is still poorly understood.
The aim of this study is to investigate the drought response mechanism and identify key regulatory genes in S. psammophila. In order to achieve this, physiological responses to drought stress were analyzed. RNA-Sequencing technology was used to identify differentially expressed genes (DEGs) of root tissue of S. psammophila at different time points (0, 6, 24 and 96 h) after drought treatment. Enrichment analysis was performed to identify pivotal functional categories involved in drought response. Based on the DEGs, gene co-expression network was constructed to identify hub genes associated with drought tolerance, and two hub genes were selected for functional verification through overexpression in A. thaliana. Our results contribute to a better understanding of drought response in S. psammophila; and suggest that co-expression network has important potential to facilitate the identification of critical genes involved in drought tolerance.
Physiological responses to drought stress in S. psammophila
RNA-Sequencing profiling and global transcriptome changes during drought stress in S. psammophila
Approximately 83.50% of total reads were aligned to S. purpurea reference genome version 1.0, a well assembled Salix genome until now (Additional file 1). Principal component analysis (PCA) was carried out on gene expression dataset to confirm the uniformity between biological replicates and investigate the relationships of samples at the four stages (Fig. 2b). Tight cluster between two biological replicates indicated the reliability of our RNA-Sequencing dataset, while greatest distinctness existed between 6 h samples and control (0 h) samples, with PC1 and PC2 explaining 89.9% of the total variance. Hierarchical clustering analysis supported these results (Fig. 2c).
DEGs were identified in different comparisons, which were classified into two groups: Group I for different treatment timepoints compare to control (6 h/0 h, 24 h/0 h and 96 h/0 h) and Group II in time-course manner (6 h/0 h, 24 h/6 h and 96 h/24 h) (Fig. 2a). The largest DEG set was identified in comparison “6 h/0 h” (a total of 5308 DEGs including 2697 up- and 2611 down-regulated genes), suggesting that S. psammophila roots rapidly reprogrammed cellular response in transcriptomic level at early stage (6 h) under drought stress. In contrast, the smallest DEG set was identified in comparison “24 h/6 h” (1181 DEGs including 494 up- and 687 down-regulated genes), followed by comparison “96 h/24 h” (2010 DEGs including 1485 up- and 525 down-regulated genes) (Fig. 2d). Comparative analysis revealed that the overlapped DEGs in three comparisons of Group I (1,198 common genes, 14.7% of total DEGs) were greater than that in Group II (207 common genes, 3.2% of total DEGs) (Fig. 2a and e). The greatest subset of stage-specific DEGs was in comparison “6 h/0 h” in both two groups (2323 DEGs in Group I and 3478 DEGs in Group II). In addition, less or no overlap of up-regulated and down-regulated genes in Group II (Fig. 2e).
Functional annotation and pathway enrichment analysis of DEGs
In MF category, the GO terms such as ‘ubiquitin-protein ligase activity’, ‘serine-type endopeptidase inhibitor activity’, ‘peptidase inhibitor activity’, ‘endopeptidase inhibitor activity’ and ‘carbohydrate binding’ were mainly enriched in Group I comparisons (6 h/0 h, 24 h/0 h and 96 h/0 h) in up-DEGs; the GO terms of ‘polysaccharide binding’, ‘pattern binding’ and ‘chitin binding’ were specifically enriched in 96 h/0 h in up-DEGs. The GO terms such as ‘glucosyltransferase activity’, ‘copper ion binding’ and ‘cellulose synthase activity’ were enriched in Group I comparisons in down-DEGs (Additional file 2: Figure S1). In CC category, up-regulated genes were significantly enriched in GO terms of ‘photosystem’, ‘thylakoid’, ‘nucleosome’, ‘microtubule’, ‘chromosome’ and ‘cell wall’ mainly in 96 h/24 h; down-regulated genes in Group I comparisons were significantly enriched in GO terms of ‘periplasmic space’, ‘microtubule’, ‘membrane’, ‘external encapsulating structure’, ‘cell wall’, ‘cell envelope’, etc. (Additional file 2: Figure S2).
To gain further insight into the molecular response to drought of S. psammophila in pathway level, we performed an enrichment analysis based on the KEGG pathways. A total of 15 pathways were significantly enriched in this process (Additional file 2: Figure S3 and Additional file 4). In particular, pathways of ‘phenypropanoid biosynthesis’, ‘metabolism of xenobiotics by cytochrome P450’, ‘phenlylalanine metabolism’ and ‘glutathione metabolism’ were significantly enriched in most comparisons, and the degree of enrichment increased with time extension under drought treatment. The pathway of ‘nitrogen metabolism’ was significantly enriched in early and medium stages (6 h/0 h and 24 h/0 h); the pathway of ‘starch and sucrose metabolism’ was significantly enriched in early and late stages (6 h/0 h and 96 h/0 h); the pathway of ‘pentose and glucuronate interconversions’ was significantly enriched in medium and late stages (24 h/0 h and 96 h/0 h). These three pathways also prominent in comparisons 24 h/6 h and 96 h/24 h. Moreover, the pathways of ‘flavonoid biosynthesis’, ‘cell cycle’, ‘gap junction’, ‘tyrosine metabolism’, ‘cyanoamino acid metabolism’ and ‘alpha-linolenic acid metabolism’ were specifically enriched in different stages in Group I comparisons (Additional file 2: Figure S3).
K-means clustering of DEGs
Transcription factors (TFs) act as a pivotal part of transcriptional regulation in plant development and stress response. We identified 672 TFs from 45 gene families were differentially expressed in the nine clusters (Additional file 2: Figure S4). The largest number of TFs were distributed in cluster 3 (164 TFs) and cluster 2 (109 TFs). TF families WRKY, NAC, MYB and ERF were abundant in these two clusters representing genes that were quickly induced under the early stage during drought stress (Fig. 4). Except for them, the members of LBD, HD-ZIP, bZIP, GRAS, G2-like, C2H2 and bHLH families were widely and unevenly distributed among the nine clusters (Additional file 2: Figure S4).
Construction of co-expression network and identification of hub genes
Overexpression of hub genes SpMDP1 or SpWRKY33 confers drought tolerance in Arabidopsis
In the co-expression network, SpMDP1 had the highest connectivity with 380 genes, most of which have previously been reported to participate in drought response or improve drought tolerance, for instance, several transcription factors (ERF, bHLH, WRKY, C2H2, bZIP, MYB, NAC, etc.), late embryogenesis abundant (LEA) hydroxyproline-rich glycoprotein (LEA-HRGP), CBL-interacting protein kinase, mitogen-activated protein kinase (Additional file 7). The Arabidopsis homologous genes of SpMDP1-co-expressed genes were searched using sequences alignment (Additional file 7). To validate the relationship of SpMDP1 and its co-expressed genes, the expression patterns of 30 randomly selected Arabidopsis homologous genes in the co-expression sub-network of SpMDP1 were compared between two-week-old SpMDP1-overexpressing plants and WT plants. Among them, the expression levels of 29 genes were unambiguously detected: 13 genes were up-regulated, four genes were down-regulated, and the expression levels of other 12 genes did not change (Additional file 2: Figure S6). Of the up-regulated genes, most genes were up-regulated by > 2-fold in transgenic plants, such as WRKY75 (2.3- to 3.2-fold), LEA-HRGP (2.1- to 3.1-fold), HSP17.6 (3.2- to 4.5-fold), GSTU25 (3.6- to 4.8-fold), and NAC90 (18- to 24-fold) (Additional file 2: Figure S6).
Accumulation of osmotic adjustment and enhancement of antioxidant enzyme activities are vital mechanisms of drought adaptation in plants [32, 33]. Osmotic adjustment through accumulating compatible solutes can promote water absorption from hypertonic solution or drought soils [34, 35]. Soluble sugars and proline are important compatible solutes, which regulate cell osmotic status, scavenge free radicals and protect cell membranes [33, 36]. Moreover, proline is proposed as a molecular chaperone functioning in stabilizing protein structure and as a component contributing to buffering of cytosolic pH [37, 38]. Reactive oxygen species (ROS) production is enhanced in cellular compartments under drought stress, and overproduction of ROS initiated uncontrolled oxidative cascades that damage cellular membranes and other cellular components and eventually cell death . Antioxidant enzymes (e.g., POD, SOD, and CAT) have the ability to scavenge ROS and maintain redox equilibrium [40, 41]. Accumulation of osmotic adjustment and increase of scavenging enzyme activity have been detected during drought in wheat, maize, soybean, rice, potato, etc. [33, 39]. In our study, the findings of accumulation of soluble sugars and proline and enhancement of antioxidant enzyme activities suggested S. psammophila might resist drought through osmotic adjustment and ROS homeostasis to provide protection from cellular damage.
Besides of morphological and physiological adaptation, extensive transcriptomic reprogramming occurred to strive for survival under drought condition [42, 43]. Thus, understanding molecular mechanisms of drought response of S. psammophila is crucial to breed drought-tolerance willows. Based on the RNA-Sequencing profiling, we identified 8172 DEGs during drought stress in roots of S. psammophila, most of which enriched in multiple stress-related processes including transcriptional regulation, response to various stresses, cell death, protein phosphorylation and ubiquitination, ncRNA and DNA metabolic, etc. Thereinto, some processes have been confirmed as essential regulators or associated with plant tolerance to drought or other osmotic stresses. For instance, transcriptional regulation has been identified as an important regulatory mechanism of transcriptomic reprogramming in response to drought . Transcriptional regulation was enriched in both clusters 2 and 3 (Fig. 4), and largest number of transcription factors existed in these two clusters (Additional file 2: Figure S4), suggesting this mechanism was dramatically evoked and executed function after short-term drought stress. Phosphoproteins have been identified under drought stress in several plants, and some proteins have been demonstrated to regulate drought tolerance via mediated phosphorylation [44, 45]. Furthermore, temporal patterns of the enriched GO terms showed that these biological processes occurred according to the time stages. These results revealed that comprehensive and well-ordered biological events motivated S. psammophila to resist and adapt to drought stress.
Based on whole-transcriptome datasets, WGCNA provides an effective approach to identify key regulators and it is widely used in various studies to explain molecular mechanisms of complex traits, such as growth, yield and stress-resistance. For instance, two potential key drivers for seed trait formation, GA20OX and NFYA, have been screened from co-expression network in soybean . WRKY40 and copper transport protein have been identified to be the hub genes involved in defense against whitefly infestation in cotton . Three major hubs including MYB2, phosphoglycerate kinase and heat shock protein have been identified during salinity and drought stresses in rice . In our study, we identified eight hub genes (MDP-1, RGLG2, Mod(r) protein, CML8, RLCK, GRAM domain-containing protein, WRKY33, and an unknown gene) that were co-expressed with a large number of genes in S. psammophila, suggesting their key positions in drought stress response. Through functional validation, two hub genes SpMDP1 and SpWRKY33 were proved to participate in plant drought tolerance. Thus, our study indicated that identification of candidate drought tolerance genes through co-expression network was highly efficient and credible.
As a member of haloacid dehalogenase superfamily, MDP1 contains a conserved motif (Asp-X-Asp-X-Thr) , and its specific function in abiotic stress tolerance in plants remains unknown until now. In our study, SpMDP1 had the highest connectivity with 380 genes, and its overexpression transgenic Arabidopsis could improve the tolerance to drought. This result provides new progress and evidence for MDP1 involved in plant drought tolerance. Glycosylation is an important posttranslational modification of proteins, and it plays a crucial role in impacting protein functions and regulating biological processes . For example, the N-glycosylation occurs in the proteins mainly involved in cell wall metabolism, stress response and proteolysis under drought in common bean . Studies have demonstrated that MDP1 can cooperate with fructosamine-3-kinase to free proteins from the glycation products derived from Glu-6-P, indicating that MDP1 may be the key component for glycosylation-involved protein repair [51, 52]. Whether the drought tolerance improved by SpMDP1 is mediated by protein repair is needed to further research.
WRKY33 belongs to the evolutionarily conserved WRKY transcription factor. The expression of WRKY33 is induced by pathogens and environmental stimuli, and its function in plant tolerance stresses has been confirmed through gene silencing and complementation [53, 54]. In A. thaliana, atwrky33 mutants are compromised not only in plant tolerance to necrotrophic pathogens and Botrytis cinerea infection, but also in tolerance to salinity and heat stresses [53, 55, 56]. Complementation of Arabidopsis atwrky33 mutant plants by tomato (Solanum lycopersicum) SlWRKY33A and SlWRKY33B can fully restore resistance to Botrytis and heat tolerance . In this study, we demonstrated that overexpression of the hub gene SpWRKY33 could enhance the tolerance to drought stress in transgenic Arabidopsis plants. Recent study has indicated that AtWRKY33 interacts with two VQ proteins, SIGMA FACTOR-INTERACTING PROTEIN1 (SIB1) and SIB2, which act as activators of AtWRKY33 in plant defence against necrotrophic pathogens . Our study also identified a VQ motif-containing protein (SapurV1A.0943 s0020) was co-expressed with SpWRKY33 in the co-expression network (Additional file 8), and its expression level was up-regulated in SpWRKY33-overexpressing plants (Additional file 2: Figure S7). Moreover, the analysis of genome-wide transcriptional responses to drought in two willow genotypes revealed that WRKY33 has been identified as one of candidate genes with a putative function in drought response . These results indicate that SpWRKY33 is an ideal gene used on genetic engineering for the creation of new varieties with high tolerance to stresses in plants.
Except SpMDP1 and SpWRKY33 which were performed functional validation in this study, the other six hub genes might also participate in drought tolerance in plants. For example, AtRGLG2 negatively regulates the drought stress response by mediating AtERF53 transcriptional activity in A. thaliana . The expression level of AtCML8 is strongly and transiently induced by Pseudomonas syringae and overexpression of AtCML8 confers on plants a better resistance to pathogenic bacteria compared to WT, suggesting that AtCML8 plays a role in plant immunity . Although there are no reports of function of CML8 in drought tolerance; other members of the CML family, such as Oryza sativa OsCML4 and Solanum habrochaites ShCML44, have been confirmed to enhance tolerance to drought stress in plants [61, 62]. Further studies are needed to investigate whether these hub genes contribute to drought tolerance in S. psammophila.
In this study, we analyzed DEGs in S. psammophila roots under a time-course drought stresses, performed analysis of biological processes and pathways, and identified eight hub genes based on the co-expression network. We found the accumulation of osmotic adjustment substances and enhancement of antioxidant enzyme activities under drought stress in S. psammophila. Abundant DEGs were enriched in several stress-related processes including transcriptional regulation, response to various stresses, cell death, etc. Through overexpression in A. thaliana, two hub genes, SpMDP1 and SpWRKY33, have been confirmed to enhance the drought tolerance in transgenic plants. Our results will contribute to parse the mechanism of drought tolerance of S. psammophila and facilitate identification of critical genes involved in drought tolerance for breeding willows.
Cultivation of S. psammophila and drought stress treatment
In our study, S. psammophila materials were obtained from the germplasm collection base of S. psammophila (E 110°38′59.1″, N 40°14′15.5″) in Ordos Dalad, Inner Mongolia. One-year-old twigs of S. psammophila were cut into 15 cm cuttings, water cultured in 1/2 strength Hoagland’s solution, and grown in a growth chamber under long-day conditions with 16 h light and 8 h dark at 24–25 °C. Ten-week-old seedlings were subjected to 22% (w/v) polyethylene glycol 6000 (PEG 6000)-simulated drought stress for four time points (0, 6, 24, and 96 h). At the end of each time point, fine roots were harvested, frozen immediately in liquid nitrogen, and stored in − 80 °C refrigerator for further analysis and sequencing. We performed three biological replicates.
Measurement of physiological parameters of S. psammophila under drought stress
Electric conductivity (EC) of S. psammophila roots was measured by conductivity meter (DDS-307). Relative water content (RWC) was calculated as previous study . The proline content was measured with ninhydrin method . The soluble sugar content was measured with anthrone method . For enzyme assays, superoxide dismutase (SOD) activity was determined by measuring its ability to inhibit the photochemical reduction of nitroblue tetrazolium at 560 nm , peroxidase (POD) activity was measured based on the change in absorbance of 470 nm due to guaiacol oxidation , catalase (CAT) activity was determined by monitoring the disappearance of H2O2 . All experiments were performed three technical replicates, and the data were analyzed with Student’s t test.
RNA isolation, library preparation, Illumina sequencing
Total RNA was extracted from roots using the RNeasy Plant Mini Kit (Qiagen, Germany) based on the manufacturer’s protocol. The RNA quality and quantity were detected by the NanoDrop system and Agilent2100 Bioanalyzer. The cDNA library was constructed and sequenced using Illumina HiSeq2500 platform by Biomarker Technologies Corporation (Beijing, China), with paired-end sequencing and read lengths of 125 bp. After quality-control procedure with filtering adapter sequences and low-quality reads, the clean data were obtained for subsequent analysis.
Mapping and quantification of RNA-Sequencing data
We selected S. purpurea genome version 1.0 from Phytozome (https://phytozome.jgi.doe.gov/pz/portal.html#!info?alias=Org_Spurpurea) as reference genome, which is 392 Mb in size with 37,865 protein-coding loci and 61,520 protein coding transcripts. Clean reads were mapped to the reference genome with allowance for 5% mismatches using TopHat2 software . Gene expression levels were calculated as reads per kilobase of transcript sequence per million base pairs sequenced (FPKM) using Cufflinks . The differentially expressed genes (DEGs) were identified by DESeq . A threshold of false discovery rate (FDR) ≤ 0.01 and an absolute value of log2 ratio > 2 were used to retrieve DEGs.
Pathway analysis and functional categorization
To study the biological function of DEGs, gene set enrichment with gene ontology (GO) terms was performed using Blast2GO . Metabolic pathways were analyzed using kyoto encyclopedia of genes and genomes (KEGG) program . Fischer’s exact test was used to assess the significance of GO categories and KEGG pathways.
Co-expression gene network generation
Quantitative real-time PCR (qRT-PCR)
Primer 3 software was used to design the primers, and qRT-PCR was performed according to the previous described method . SpUBC (Ubiquitin-conjugating enzyme E2) and AtActin were used as the reference genes in S. psammophila  and A. thaliana, respectively. Four technical replicates and three biological replicates of each sample were performed. All of the primer sequences were listed in Additional file 9. For homologous genes identification between Arabidopsis and Salix, the BLASTp best hit of Salix genes from the co-expression network in Arabidopsis genome were used for qRT-PCR analysis in Arabidopsis transgenic lines.
Plasmid construction and transformation in A. thaliana
The full length CDS sequences of SpMDP1 and SpWRKY33 were independently amplified from S. psammophila and cloned into pDONR222.1 for sequencing, and the correct coding sequences were sub-cloned into pMDC32 under driven of the cauliflower mosaic virus (CaMV) 35S promoter using Gateway system, respectively. Both above constructs were transferred into A. thaliana by floral dip method . After screened on medium containing 25 mg·L− 1 hygromycin, more than 30 transgenic lines were obtained and the expression levels of these two target genes were detected using qRT-PCR.
Drought tolerance analysis of overexpression plants
WT and homozygous T3 generation transgenic Arabidopsis lines were used in drought tolerance experiments. One-week-old sterilized seedlings were transferred into 1/2 strength MS medium containing 0 mmol/L (control) and 200 mmol/L mannitol (simulated drought stress), respectively. Ten days later, phenotypic traits were compared, and the root length and fresh weight of seedlings were counted. Drought treatment was also applied to three-week-old seedlings in soil with sufficient water by withholding watering for 14 days. Five biological replicates were performed to ensure results’ reliability, and data were analyzed by t test.
We would like to thank the reviewers and editors for their careful reading and helpful comments on this manuscript.
HJ, JZ, JL and JH conceived and designed the research; JL, PS and YZ performed all experiments; HJ and JZ conducted data analysis; HJ, JZ and JL wrote the manuscript; XX, ML and JH contributed with valuable discussions. All authors read and approved the final manuscript.
This work was supported by National Nonprofit Institute Research Grant of CAF (CAFYBB2017ZY008), (CAFYBB2018ZY001–9), and Natural Science Foundation of China (31800570). The founders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Ethics approval and consent to participate
Consent for publication
The authors declare that they have no competing interests.
- 1.Hassan R, Scholes R, Ash N. Ecosystems and human well-being: current state and trends. Washington DC: Island Press; 2005.Google Scholar
- 3.Williams AP, Allen CD, Macalady AK, Griffin D, Woodhouse CA, Meko DM, Swetnam TW, Rauscher SA, Seager R, Grissinomayer HD. Temperature as a potent driver of regional forest drought stress and tree mortality. Nat Clim Chang. 2013;3(3):292–7.Google Scholar
- 4.Dai A. Increasing drought under global warming in observations and models. Nat Clim Chang. 2013;3(1):52–8.Google Scholar
- 10.Lenka SK, Katiyar A, Chinnusamy V, Bansal KC. Comparative analysis of drought-responsive transcriptome in Indica rice genotypes with contrasting drought tolerance. Plant Biotechnol J. 2015;9(3):315–27.Google Scholar
- 11.Fei L, Zhang H, Zhao H, Gao T, Song A, Jiang J, Chen F, Chen S. Chrysanthemum CmHSFA4 gene positively regulates salt stress tolerance in transgenic chrysanthemum. Plant Biotechnol J. 2018;16(7):1311–21.Google Scholar
- 23.Kuzovkina YA, Weih M, Romero MA, Charles J, Hust S, Mcivor I, Karp A, Trybush S, Labrecque M, Teodorescu TI. Salix: botany and global horticulture. Hortic Rev. 2008;34(1):447–89.Google Scholar
- 25.Bonosi L, Ghelardini L, Weih M. Growth responses of 15 Salix genotypes to temporary water stress are different from the responses to permanent water shortage. Trees. 2010;24(5):843–54.Google Scholar
- 27.Huang J, Zhou Y, Yin L, Wenninger J, Zhang J, Hou G, Zhang E, Uhlenbrook S. Climatic controls on sap flow dynamics and used water sources of Salix psammophila in a semi-arid environment in Northwest China. Environ Earth Sci. 2015;73(1):289–301.Google Scholar
- 28.Li CJ, Yang X, Zhang Z, Zhou D, Zhang L, Zhang SC, Chen JM. Hydrothermal liquefaction of desert shrub Salix psammophila to high value-added chemicals and hydrochar with recycled processing water. Bioresources. 2013;8(2):2981–97.Google Scholar
- 29.Yang X, Lyu H, Chen K, Zhu X, Zhang S, Chen J. Selective extraction of bio-oil from hydrothermal liquefaction of Salix psammophila by organic solvents with different polarities through multistep extraction separation. Bioresources. 2014;9(3):5219–33.Google Scholar
- 32.Hsiao TC, Acevedo E, Fereres E, Henderson DW. Stress metabolism: water stress, growth, and osmotic adjustment. Philos Trans R Soc Lond Ser B Biol Sci. 1976;273:479–500.Google Scholar
- 33.Zivcak M, Brestic M, Sytar O. Osmotic adjustment and plant adaptation to drought stress. In: Hossain MA, Wani SH, Bhattacharjee S, Burritt DJ, Tran LP, editors. Drought stress tolerance in plants. Switzerland: Springer International Publishing; 2016. p. 105–143.Google Scholar
- 34.Chaves MM, Maroco J, Pereira JS. Understanding plant responses to drought-from genes to the whole plant. Funct Plant Biol. 2003;30(3):239–64.Google Scholar
- 35.Liu C, Liu Y, Guo K, Fan D, Li G, Zheng Y, Yu L, Yang R. Effect of drought on pigments, osmotic adjustment and antioxidant enzymes in six woody plant species in karst habitats of southwestern China. Environ Exp Bot. 2011;71(2):174–83.Google Scholar
- 38.Filippou P, Bouchagier P, Skotti E, Fotopoulos V. Proline and reactive oxygen/nitrogen species metabolism is involved in the tolerant response of the invasive plant species Ailanthus altissima to drought and salinity. Environ Exp Bot. 2014;97:1–10.Google Scholar
- 39.De Carvalho MHC. Drought stress and reactive oxygen species: production, scavenging and signaling. Plant Signal Behav. 2008;3(3):156–65.Google Scholar
- 42.Bechtold U, Penfold CA, Jenkins DJ, Legaie R, Moore JD, Lawson T, Matthews JS, Vialet-Chabrand SR, Baxter L, Subramaniam S. Time-series transcriptomics reveals that AGAMOUS-LIKE22 links primary metabolism to developmental processes in drought-stressed Arabidopsis. Plant Cell. 2016;28(2):345–55.PubMedPubMedCentralGoogle Scholar
- 44.Zhang M, Lv D, Ge P, Bian Y, Chen G, Zhu G, Li X, Yan Y. Phosphoproteome analysis reveals new drought response and defense mechanisms of seedling leaves in bread wheat (Triticum aestivum L.). J Proteome. 2014;109:290–308.Google Scholar
- 52.Chondrogianni N, Petropoulos I, Grimm S, Georgila K, Catalgol B, Friguet B, Grune T, Gonos ES. Protein damage, repair and proteolysis. Mol Asp Med. 2014;35:1–71.Google Scholar
- 53.Zheng Z, Qamar SA, Chen Z, Mengiste T. Arabidopsis WRKY33 transcription factor is required for resistance to necrotrophic fungal pathogens. Plant J. 2010;48(4):592–605.Google Scholar
- 54.Lippok B, Birkenbihl RP, Rivory G, Brummer J, Schmelzer E, Logemann E, Somssich IE. Expression of AtWRKY33 encoding a pathogen- or PAMP-responsive WRKY transcription factor is regulated by a composite dna motif containing W box elements. Mol Plant-Microbe Interact. 2007;20(4):420–9.PubMedGoogle Scholar
- 61.Yin XM, Huang LF, Zhang X, Wang ML, Xu GY, Xia X. OsCML4 improves drought tolerance through scavenging of reactive oxygen species in rice. J Plant Biol. 2015;58(1):68–73.Google Scholar
- 62.Shoaib M, Liu H, Xing Y, Saddam H, Ouyang B, Zhang Y, Li H, Ye Z. Overexpression of calmodulin-like (ShCML44) stress-responsive gene from Solanum habrochaites enhances tolerance to multiple abiotic stresses. Sci Rep. 2016;6:31772.Google Scholar
- 64.Bates LS, Waldren RP, Teare ID. Rapid determination of free proline for water-stress studies. Plant Soil. 1973;39(1):205–7.Google Scholar
- 65.Dubois M, Gilles KA, Hamilton JK, Rebers PA, Smith F. Colorimetric method for determination of sugars and related substances. Anal Chem. 1956;28(3):350–6.Google Scholar
- 69.Kim D, Pertea G, Trapnell C, Pimentel H, Kelley R, Salzberg SL. TopHat2: accurate alignment of transcriptomes in the presence of insertions, deletions and gene fusions. Genome Biol. 2013;14(4):1–13.Google Scholar
- 70.Trapnell C, Williams BA, Pertea G, Mortazavi A, Kwan G, Van Baren MJ, Salzberg SL, Wold BJ, Pachter L. Transcript assembly and quantification by RNA-Seq reveals unannotated transcripts and isoform switching during cell differentiation. Nat Biotechnol. 2010;28(5):511–5.PubMedPubMedCentralGoogle Scholar
- 71.Anders S, Huber W. Differential expression analysis for sequence count data. Genome Biol. 2010;11(10):1–12.Google Scholar
- 73.Kanehisa M, Goto S. KEGG: Kyoto encyclopedia of genes and genomes. Nucleic Acids Res. 1999;28(1):27–30.Google Scholar
- 74.Langfelder P, Horvath S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinf. 2008;9(1):559.Google Scholar
- 76.Jia H, Li J, Zhang J, Ren Y, Hu J, Lu M. Genome-wide survey and expression analysis of the stress-associated protein gene family in desert poplar, Populus euphratica. Tree Genet Genomes. 2016;12(4):78.Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.